JP7340569B2 - Predicted image correction device, image encoding device, image decoding device, and program - Google Patents

Description

The present invention relates to a predictive image correction device, an image encoding device, an image decoding device, and a program.

In video encoding technology, efficient encoding (compression) is achieved by dividing a current image in units of frames (pictures) and performing conversion, prediction, etc. in units of blocks obtained. There are two types of prediction methods: intra prediction and inter prediction.

Intra prediction is a method of generating a predicted image by referring to decoded adjacent blocks adjacent to a block to be encoded or decoded (hereinafter referred to as "target image block"). Inter prediction is a method of generating a predicted image by referring to a decoded frame different from the current frame to which the target image block belongs.

Patent Document 1 discloses that a target image block is predicted by inter prediction to generate an inter predicted image, a target image block is predicted by intra prediction to generate an intra predicted image, and the inter predicted image and the intra predicted image are weighted. A predictive image correction device that averages is described.

Specifically, the predicted image correction device described in Patent Document 1 evaluates the continuity of an inter predicted image corresponding to a target image block and a decoded adjacent block adjacent to the target image block, and determines whether the inter predicted image is discontinuous or not. If it is evaluated as , the end region of the inter-predicted image is corrected by filter processing using the decoded adjacent blocks.

US Patent Application Publication No. 2013/051467

The predicted image correction device described in Patent Document 1 applies filter processing according to the evaluation of continuity between an inter predicted image and a decoded adjacent block, regardless of the prediction accuracy of inter prediction.

Therefore, if there is an object boundary between the inter-predicted image and the decoded adjacent block, filter processing is applied to the edge area even if the prediction accuracy of the edge area of the inter-predicted image is high. Therefore, there is a problem that the accuracy of the predicted image may decrease due to the filtering process.

Therefore, an object of the present invention is to provide a predicted image correction device, an image encoding device, an image decoding device, and a program that can appropriately correct a predicted image.

The predicted image correction device according to the present invention includes an inter prediction unit that predicts a target image block obtained by dividing an image in units of frames by inter prediction, and generates an inter prediction block corresponding to the target image block; an intra prediction unit that predicts a target image block by intra prediction and generates an intra prediction block corresponding to the target image block; and a predicted image that performs a weighted average process of the inter prediction block and the intra prediction block as a correction process. a correction unit, and the predicted image correction unit adjusts the weight in the weighted average processing according to a prediction mode that controls intra prediction processing for a decoded adjacent block adjacent to the target image block. shall be. Further, the predicted image correction device according to the first feature predicts the target image block obtained by dividing the current image in units of frames using a plurality of reference images, thereby predicting the target image block corresponding to the target image block. a prediction unit that generates an image; a prediction accuracy evaluation unit that evaluates the prediction accuracy of the predicted image based on the similarity between the plurality of reference images used to generate the predicted image; and a prediction accuracy evaluation unit that is adjacent to the target image block. a correction unit that performs correction processing on the predicted image using decoded adjacent blocks, and the correction unit controls the correction processing based on at least an evaluation result by the prediction accuracy evaluation unit. shall be.

Note that prediction performed using a plurality of reference images is typically bi-prediction in inter prediction, but is not limited to this; for example, prediction performed using intraBC mode ( A similar method can be applied even when a predicted image is generated by making multiple references, such as in intra block copy mode).

The gist of the image encoding device according to the second feature is that it includes the predicted image correction device according to the first feature.

The gist of the image decoding device according to the third feature is that it includes the predictive image correction device according to the first feature.

The gist of the program according to the fourth feature is to cause a computer to function as the predicted image correction device according to the first feature.

According to the present invention, it is possible to provide a predicted image correction device, an image encoding device, an image decoding device, and a program that can appropriately correct a predicted image.

FIG. 1 is a diagram showing the configuration of an image encoding device according to an embodiment. 1 is a diagram showing the configuration of an image decoding device according to an embodiment. It is a figure showing an example of inter prediction. FIG. 3 is a diagram showing an example of a predicted image generated by inter prediction. It is a figure showing an example of composition of a prediction accuracy evaluation part concerning an embodiment. FIG. 3 is a diagram illustrating an example of a configuration of a predicted image correction section according to an embodiment. It is a figure showing an example of operation of a continuity evaluation part concerning an embodiment. FIG. 3 is a diagram illustrating an example of a predicted image correction operation according to the embodiment. It is a figure which shows the structure of the predicted image correction part based on the example of a change of embodiment. It is a figure which shows the example of operation of the continuity evaluation part and filter processing part based on the example of a change of embodiment.

An image encoding device and an image decoding device according to embodiments will be described with reference to the drawings. The image encoding device and the image decoding device according to the embodiment encode and decode moving images represented by MPEG. In the description of the drawings below, the same or similar parts are designated by the same or similar symbols.

(1. Configuration of image encoding device)
FIG. 1 is a diagram showing the configuration of an image encoding device 1 according to this embodiment. As shown in FIG. 1, the image encoding device 1 includes a block division section 100, a subtraction section 101, a transformation section 102a, a quantization section 102b, an entropy encoding section 103, an inverse quantization section 104a, It includes an inverse transformation section 104b, a synthesis section 105, a memory 106, an intra prediction section 107, an inter prediction section 108, a prediction accuracy evaluation section 109, and a predicted image correction section (correction section) 110. In this embodiment, the intra prediction unit 107, the inter prediction unit 108, the prediction accuracy evaluation unit 109, and the predicted image correction unit 110 constitute a predicted image correction device.

The block dividing unit 100 divides an input image in frame (or picture) units into block-shaped small areas, and outputs the image blocks to the subtraction unit 101 (and the inter prediction unit 108). The size of the image block is, for example, 32×32 pixels, 16×16 pixels, 8×8 pixels, or 4×4 pixels. An image block is a unit for encoding by the image encoding device 1 and a unit for decoding by the image decoding device 2, and such an image block is referred to as a target image block. Note that the shape of the image block is not limited to a square, but may be a rectangle.

The subtracting unit 101 calculates a prediction residual indicating a difference in pixel units between the target image block input from the block dividing unit 100 and a predicted image (predicted image block) corresponding to the target image block. Specifically, the subtraction unit 101 calculates a prediction residual by subtracting each pixel value of the predicted image from each pixel value of the current block to be encoded, and outputs the calculated prediction residual to the conversion unit 102a. In this embodiment, the predicted image is corrected by a predicted image correction unit 110 described later, and is input from the predicted image correction unit 110 to the subtraction unit 101.

The transformation unit 102a and the quantization unit 102b constitute a transformation/quantization unit 102 that performs orthogonal transformation processing and quantization processing on a block-by-block basis.

The transformation unit 102a performs orthogonal transformation on the prediction residual input from the subtraction unit 101 to calculate transformation coefficients, and outputs the calculated transformation coefficients to the quantization unit 102b. The orthogonal transformation refers to, for example, a discrete cosine transform (DCT), a discrete sine transform (DST), a KarhunenLoeve transform (KLT), or the like.

The quantization unit 102b quantizes the transformation coefficients input from the transformation unit 102a using a quantization parameter (Qp) and a quantization matrix to generate quantized transformation coefficients. The quantization parameter (Qp) is a parameter commonly applied to each transform coefficient within a block, and is a parameter that determines the coarseness of quantization. The quantization matrix is a matrix having as elements quantization values when quantizing each transform coefficient. The quantization unit 102b outputs quantization control information, generated quantization transform coefficient information, etc. to the entropy encoding unit 103 and the inverse quantization unit 104a.

The entropy encoding unit 103 performs entropy encoding on the quantized transform coefficients input from the quantization unit 102b, performs data compression to generate encoded data (bitstream), and converts the encoded data into an image code. output to the outside of the conversion device 1. For entropy encoding, a Huffman code, CABAC (Context-based Adaptive Binary Arithmetic Coding), or the like can be used. Note that information regarding prediction is input to the entropy encoding unit 103 from the intra prediction unit 107 and the inter prediction unit 108. The entropy encoding unit 103 also performs entropy encoding of this information.

The inverse quantization unit 104a and the inverse transformation unit 104b constitute an inverse quantization/inverse transformation unit 104 that performs inverse quantization processing and inverse orthogonal transformation processing on a block-by-block basis.

The dequantization unit 104a performs dequantization processing corresponding to the quantization processing performed by the quantization unit 102b. Specifically, the dequantization unit 104a restores the transformation coefficients by dequantizing the quantized transformation coefficients input from the quantization unit 102b using a quantization parameter (Qp) and a quantization matrix. Then, the restored transform coefficients are output to the inverse transform section 104b.

The inverse transform unit 104b performs an inverse orthogonal transform process corresponding to the orthogonal transform process performed by the transform unit 102a. For example, if the transform unit 102a performs a discrete cosine transform, the inverse transform unit 104b performs an inverse discrete cosine transform. The inverse transform unit 104b performs inverse orthogonal transform on the transform coefficients input from the inverse quantization unit 104a to restore the prediction residual, and outputs the restored prediction residual, which is the restored prediction residual, to the synthesis unit 105. do.

The combining unit 105 combines the restored prediction residual inputted from the inverse transformation unit 104b with the predicted image inputted from the predicted image correction unit 110, pixel by pixel. The synthesis unit 105 reconstructs (decodes) the target image block by adding each pixel value of the restored prediction residual and each pixel value of the predicted image, and stores the reconstructed image block, which is the reconstructed target image block, in the memory 106. Output to. Note that such a reconstructed image block is sometimes referred to as a decoded block.

The memory 106 stores the reconstructed image block input from the synthesis unit 105. Memory 106 stores reconstructed image blocks in units of frames.

The intra prediction unit 107 generates an intra predicted image by referring to decoded adjacent blocks adjacent to the target image block among the reconstructed image blocks (decoded blocks) stored in the memory 106 . The intra prediction unit 107 selects an optimal intra prediction mode and performs intra prediction using the selected intra prediction mode. A plurality of intra prediction modes corresponding to a plurality of intra prediction directions are defined in advance. The intra prediction direction refers to the direction of adjacent reference pixels with respect to the target pixel when predicting the target pixel in the target image block with reference to the adjacent pixels adjacent to the target image block. That is, the intra prediction mode (intra prediction direction) determines adjacent reference pixels to be used for prediction of each pixel in the target image block. The intra prediction unit 107 outputs the intra predicted image (and/or adjacent reference pixels) to the predicted image correction unit 110 and also outputs information on the selected intra prediction mode to the entropy encoding unit 103.

The inter prediction unit 108 performs inter prediction to predict the target image block using the frame-by-frame reconstructed image (decoded image) stored in the memory 106 as a reference image. Specifically, the inter prediction unit 108 calculates a motion vector using a technique such as block matching, and generates an inter predicted image based on the motion vector. The inter prediction unit 108 selects the optimal inter prediction method from among inter prediction using multiple reference images (typically, bi-prediction) and inter prediction using one reference image (unidirectional prediction), Perform inter prediction using the selected inter prediction method. In this embodiment, a case will be mainly described in which the inter prediction unit 108 uses bi-prediction, and the inter prediction unit 108 corresponds to a prediction unit that performs prediction using a plurality of reference images. The inter prediction unit 108 outputs the generated inter prediction image to the predicted image correction unit 110 and also outputs information regarding the selected inter prediction method and motion vector to the entropy encoding unit 103. Further, the inter prediction unit 108 outputs a plurality of reference images used for inter prediction to the prediction accuracy evaluation unit 109.

The prediction accuracy evaluation unit 109 evaluates the prediction accuracy in the edge region of the inter-predicted image based on the degree of similarity between a plurality of reference images used to generate the inter-predicted image. Specifically, when the inter prediction unit 108 performs inter prediction using a plurality of reference images, the degree of similarity between the plurality of reference images used for inter prediction is calculated for each image portion consisting of one or more pixels. , the prediction accuracy of the inter-predicted image is evaluated for each image part based on the similarity, and information on the evaluation result is output to the predicted image correction unit 110. In this embodiment, an example will be mainly described in which the prediction accuracy evaluation unit 109 calculates the similarity between a plurality of reference images used for prediction on a pixel basis and evaluates the prediction accuracy of an inter-predicted image on a pixel basis. Details of the prediction accuracy evaluation unit 109 will be described later.

The predicted image correction unit 110 corrects the inter predicted image input from the inter prediction unit 108 and the intra predicted image input from the intra prediction unit 107 by performing a weighted average. Specifically, the predicted image correction unit 110 uses a decoded adjacent block (adjacent reference pixel) corresponding to the intra-predicted image to correct the end region of the inter-predicted image by filtering. Further, the predicted image correction unit 110 controls filter processing based at least on the evaluation result by the prediction accuracy evaluation unit 109. "Controlling filter processing" includes control of whether to perform filter processing and control of filter strength. In this embodiment, an example in which the predicted image correction unit 110 controls whether or not to perform filter processing based at least on the evaluation result by the prediction accuracy evaluation unit 109 will be mainly described. For example, the predicted image correction unit 110 performs filter processing when the prediction accuracy evaluated by the prediction accuracy evaluation unit 109 is less than or equal to the first threshold, and the prediction accuracy evaluated by the prediction accuracy evaluation unit 109 is set to the first threshold. If the number exceeds , no filtering is performed. The predicted image correction unit 110 outputs the filtered inter predicted image as a predicted image to the subtraction unit 101 and the synthesis unit 105 when performing filter processing, and predicts the inter predicted image as it is when not performing filter processing. It is output as an image to the subtraction unit 101 and the composition unit 105. Note that "prediction accuracy" may be expressed as "inaccuracy of prediction." Regarding "unaccuracy of prediction," for example, the fact that the prediction accuracy is less than or equal to the first threshold value may mean that the prediction accuracy is high. Exceeding the first threshold may mean that the prediction accuracy is low. Details of the predicted image correction unit 110 will be described later.

(2. Configuration of image decoding device)
FIG. 2 is a diagram showing the configuration of the image decoding device 2 according to this embodiment. As shown in FIG. 2, the image decoding device 2 includes an entropy decoding section 200, an inverse quantization section 201a, an inverse transformation section 201b, a combining section 202, a memory 203, an intra prediction section 204, and an inter prediction section. 205, a prediction accuracy evaluation section 206, and a predicted image correction section 207. In this embodiment, the intra prediction unit 204, the inter prediction unit 205, the prediction accuracy evaluation unit 206, and the predicted image correction unit 207 constitute a predicted image correction device.

Entropy decoding section 200 decodes the encoded data generated by encoding device 1, and outputs quantized transform coefficients to dequantization section 201a. Further, the entropy decoding unit 200 decodes the encoded data, obtains information regarding prediction (intra prediction and inter prediction), and outputs information regarding prediction to the intra prediction unit 204 and the inter prediction unit 205.

The inverse quantization unit 201a and the inverse transformation unit 201b constitute an inverse quantization/inverse transformation unit 201 that performs inverse quantization processing and inverse orthogonal transformation processing on a block-by-block basis.

The dequantization unit 201a performs dequantization processing corresponding to the quantization processing performed by the quantization unit 102b of the image encoding device 1. The dequantization unit 201a restores the transform coefficients by dequantizing the quantized transform coefficients input from the entropy decoding unit 200 using a quantization parameter (Qp) and a quantization matrix, and restores the restored transform coefficients. The coefficients are output to the inverse transform section 201b.

The inverse transform unit 201b performs an inverse orthogonal transform process corresponding to the orthogonal transform process performed by the transform unit 102a of the image encoding device 1. The inverse transform unit 201b performs inverse orthogonal transform on the transform coefficients input from the inverse quantization unit 201a to restore the prediction residual, and outputs the restored prediction residual (restored prediction residual) to the synthesis unit 202. do.

The synthesis unit 202 reconstructs (decodes) the original target image block by synthesizing the prediction residual input from the inverse transformation unit 201b and the prediction image input from the prediction image correction unit 207 pixel by pixel. Then, the reconstructed image block is output to the memory 203.

The memory 203 stores the reconstructed image block input from the combining unit 202. The memory 203 stores reconstructed image blocks in units of frames. The memory 203 outputs reconstructed images (decoded images) in frame units to the outside of the image decoding device 2 in display order.

The intra prediction unit 204 generates an intra predicted image by referring to the reconstructed image block stored in the memory 203 and performing intra prediction according to the intra prediction information (intra prediction mode) input from the entropy decoding unit 200. Specifically, the intra prediction unit 204 generates an intra prediction image by referring to adjacent reference pixels determined according to the intra prediction mode among the reconstructed image blocks (decoded blocks) stored in the memory 203. . The intra prediction unit 204 outputs the intra prediction image (and/or adjacent reference pixels) to the predicted image correction unit 207.

The inter prediction unit 205 performs inter prediction to predict the target image block using the frame-by-frame reconstructed image (decoded image) stored in the memory 106 as a reference image. The inter prediction unit 205 generates an inter prediction image by performing inter prediction according to the inter prediction information (motion vector information, etc.) input from the entropy decoding unit 200, and outputs the inter prediction image to the prediction image correction unit 207. Further, the inter prediction unit 205 outputs a plurality of reference images used for inter prediction to the prediction accuracy evaluation unit 206.

The prediction accuracy evaluation unit 206 performs the same operation as the prediction accuracy evaluation unit 109 of the image encoding device 1. When the inter prediction unit 205 performs inter prediction using a plurality of reference images, the prediction accuracy evaluation unit 206 calculates the degree of similarity between the plurality of reference images in pixel units, thereby evaluating the prediction accuracy of the predicted image pixel by pixel. The evaluation is performed in units of units, and information on the evaluation results is output to the predicted image correction unit 207.

The predicted image correction unit 207 performs the same operation as the predicted image correction unit 110 of the image encoding device 1. The predicted image correction unit 207 uses decoded adjacent blocks (adjacent reference pixels) corresponding to the intra predicted image to correct the end regions of the inter predicted image through filter processing. The predicted image correction unit 207 performs filter processing when the prediction accuracy evaluated by the prediction accuracy evaluation unit 206 is less than or equal to the first threshold, and the prediction accuracy evaluated by the prediction accuracy evaluation unit 206 exceeds the first threshold. In this case, no filter processing is performed. The predicted image correction unit 207 outputs the filtered inter predicted image as a predicted image to the synthesis unit 202 when performing filter processing, and outputs the inter predicted image as it is as a predicted image to the synthesis unit 202 when not performing filter processing. 202. Details of the predicted image correction unit 207 will be described later.

(3. Inter prediction)
FIG. 3 is a diagram illustrating an example of inter prediction. FIG. 4 is a diagram illustrating an example of a predicted image generated by inter prediction. As a simple example of inter prediction, a case will be described in which bi-prediction used in HEVC, particularly forward and backward prediction (bidirectional prediction) is used.

As shown in FIG. 3, bi-prediction refers to frames temporally before and after the target frame (current frame). In the example of FIG. 3, prediction of a block in the t-th frame image is performed with reference to the t-1-th frame and the t+1-th frame. In motion detection, locations (blocks) similar to the target image block are detected from within the t-1 and t+1-th reference frames within a search range set by the system.

The detected location is the reference image. Information indicating the relative position of the reference image with respect to the target image block is indicated by an arrow in the figure, and is called a motion vector. The motion vector information is encoded by entropy encoding together with the frame information of the reference image in the image encoding device 1 . On the other hand, the image decoding device 2 detects a reference image based on the motion vector information generated by the image encoding device 1.

As shown in FIGS. 3 and 4, reference images 1 and 2 detected by motion detection are similar partial images that are aligned within the reference frame with respect to the target image block. The resulting image is similar to the encoding target image). In the example of FIG. 4, the target image block includes a star pattern and a partial circle pattern. Reference image 1 includes a star pattern and an overall circle pattern. Reference image 2 includes a star pattern, but does not include a circle pattern.

A predicted image is generated from these reference images 1 and 2. Note that the prediction process generally generates a predicted image having the characteristics of each reference image by averaging reference images 1 and 2 that have different but partially similar characteristics. However, the predicted image may be generated using more advanced processing, for example, signal enhancement processing using a low-pass filter, a high-pass filter, or the like. Here, since reference image 1 includes a circle pattern and reference image 2 does not include a circle pattern, when a predicted image is generated by averaging reference images 1 and 2, the circle pattern in the predicted image is The signal is halved compared to 1.

The difference between the predicted images obtained from reference images 1 and 2 and the target image block (encoding target image) is the prediction residual. In the prediction residuals shown in Figure 4, there is a large difference only in the shifted part of the edge of the star pattern and the shifted part of the circle pattern (hatched area), but other parts are predicted with good accuracy. has been performed, and the difference is reduced (in the example of FIG. 4, no difference has occurred).

The areas where no difference occurs (non-edge areas of the star pattern and background areas) are areas where the similarity between reference image 1 and reference image 2 is high, and where highly accurate predictions have been made. be. On the other hand, a portion where a large difference occurs is a portion unique to each reference image, that is, a portion where the degree of similarity between reference image 1 and reference image 2 is extremely low. Therefore, it can be seen that a portion where the degree of similarity between reference image 1 and reference image 2 is extremely low has low prediction accuracy and causes a large difference (residual difference).

In this way, when the prediction residuals containing a mixture of parts with large differences and parts with no differences are orthogonally transformed and deterioration of the transform coefficients occurs due to quantization, the deterioration of the transform coefficients is caused by inverse quantization and inverse orthogonal transform. Propagates entirely within the image (block). Then, when the target image block is reconstructed by combining the prediction residual restored by inverse quantization and inverse orthogonal transformation (restored prediction residual) with the prediction image, the non-edge part of the star pattern and the background shown in FIG. The deterioration in image quality also propagates to the parts for which highly accurate predictions have been made.

(4. Prediction accuracy evaluation department)
FIG. 5 is a diagram showing an example of the configuration of the prediction accuracy evaluation section 109 in the image encoding device 1. The prediction accuracy evaluation unit 109 evaluates the prediction accuracy of the predicted image on a pixel-by-pixel basis by calculating the similarity between a plurality of reference images used for inter prediction on a pixel-by-pixel basis. Here, an example will be described in which the prediction accuracy evaluation unit 109 evaluates the prediction accuracy for all pixels of the predicted image, but the prediction accuracy evaluation unit 109 does not necessarily need to evaluate the prediction accuracy for all pixels of the predicted image. The prediction accuracy evaluation unit 109 may evaluate the prediction accuracy of at least the pixels in the edge region of the predicted image.

As shown in FIG. 5, the prediction accuracy evaluation section 109 includes a difference calculation section (subtraction section) 109a, a normalization section 109b, and an adjustment section 109c.

The difference calculation unit 109a calculates the absolute value of the difference value between the reference image 1 and the reference image 2 for each pixel, and outputs the calculated absolute value of the difference value to the normalization unit 109b. The absolute value of the difference value is an example of a value indicating the degree of similarity. It can be said that the smaller the absolute value of the difference value, the higher the similarity, and the larger the absolute value of the difference value, the lower the similarity. The difference calculation unit 109a may perform filter processing on each reference image and then calculate the absolute value of the difference value. The difference calculation unit 109a may calculate a statistic such as a squared error, and use this statistic as the degree of similarity.

The normalization unit 109b converts the difference value of each pixel input from the difference calculation unit 109a into the absolute value of the difference value of the pixel having the maximum absolute value of the difference value within the block (that is, the absolute value of the difference value within the block). The normalized difference value, which is the absolute value of the normalized difference value, is output to the adjustment unit 109c. In this embodiment, the normalized difference value is used as a weight for weighting the restored prediction residual to be combined with the predicted image in the combining unit 105 on a pixel-by-pixel basis.

The adjustment unit 109c adjusts the normalized difference value (weight) input from the normalization unit 109b based on the quantization parameter (Qp) that determines the roughness of quantization, and outputs this weight. Since the greater the coarseness of quantization, the higher the degree of deterioration of the restored prediction residual, the adjustment unit 109c takes the degree of deterioration into account by adjusting the normalized difference value (weight) based on the quantization parameter (Qp). The restored prediction residuals can be weighted by

The estimated prediction accuracy Rij of each pixel (ij) output by the prediction accuracy evaluation unit 109 can be expressed, for example, as in equation (1) below.

Rij = 1 - (abs(Xij-Yij)/maxD x Scale(Qp))...(1)

In equation (1), Xij is the pixel value of pixel ij of reference image 1, Yij is the pixel value of pixel ij of reference image 2, and abs is a function that obtains the absolute value.

Furthermore, in equation (1), maxD is the maximum value of the difference value abs(Xij−Yij) within the block. In order to calculate maxD, it is necessary to calculate the difference values for all pixels in the block, but in order to omit this process, the maximum value of adjacent blocks that have already been processed may be used instead. If there is a value greater than the value, maxD may be normalized by clipping at the maximum value used. Alternatively, maxD may be determined from the quantization parameter (Qp) using a table that defines the correspondence between the quantization parameter (Qp) and maxD. Alternatively, a fixed value predefined in the specifications may be used as maxD.

Furthermore, in equation (1), Scale (Qp) is a coefficient multiplied according to the quantization parameter (Qp). Scale (Qp) is designed so that it approaches 1.0 when Qp is large and approaches 0 when it is small, and its degree is adjusted by the system. Alternatively, a fixed value predefined in the specifications may be used as Scale (Qp). Furthermore, in order to simplify the processing, Scale (QP) may be set to a fixed value such as 1.0 designed according to the system.

The adjustment unit 109c outputs the estimated prediction accuracy Rij. Moreover, this Rij may output weighting adjusted by a sensitivity function designed according to the system. For example, let abs(Xij-Yij)/maxD×Scale(Qp)=Rij, and not only Rij=Clip(rij, 1.0, 0.0) but also Rij=Clip(rij+offset, 1.0, 0.0 ) and an offset according to control information such as QP to adjust the sensitivity. Note that Clip (x, max, min) indicates a process of clipping at max when x exceeds max, and clipping at min when x is less than min.

The estimated prediction accuracy Rij calculated in this way has a value within the range of 0 to 1.0. Basically, estimated prediction accuracy Rij approaches 0 when the absolute value of the difference value of pixel ij between reference images is large (that is, the prediction accuracy is low), and when the absolute value of the difference value of pixel ij between reference images It approaches 1 when the value is small (that is, the prediction accuracy is high). The prediction accuracy evaluation unit 109 outputs map information consisting of the estimated prediction accuracy Rij of each pixel ij in the block to the predicted image correction unit 110 in units of blocks.

Note that the prediction accuracy evaluation unit 109 performs evaluation (calculation of estimated prediction accuracy Rij) only when applying inter prediction using multiple reference images, and performs evaluation (calculation of estimated prediction accuracy Rij) only when applying inter prediction using multiple reference images. In intra prediction processing in which there is no evaluation, it is not necessary to perform evaluation.

Furthermore, although the prediction accuracy evaluation unit 109 in the image encoding device 1 has been described, the prediction accuracy evaluation unit 206 in the image decoding device 2 is configured similarly to the prediction accuracy evaluation unit 109 in the image encoding device 1. Specifically, the prediction accuracy evaluation section 206 in the image decoding device 2 includes a difference calculation section 206a, a normalization section 206b, and an adjustment section 206c.

(5. Predicted image correction unit)
FIG. 6 is a diagram showing an example of the configuration of the predicted image correction section 110 in the image encoding device 1. As shown in FIG. 6, the predicted image correction section 110 includes a continuity evaluation section 110a, a filter determination section 110b, and a filter processing section 110c.

The continuity evaluation unit 110a evaluates the continuity between the inter predicted image input from the inter prediction unit 108 and the decoded adjacent block input from the intra prediction unit 107. The decoded adjacent block is a decoded block adjacent to the target image block. The continuity evaluation unit 110a outputs the evaluation result of continuity between the inter-predicted image and the decoded adjacent block to the filter determination unit 110b.

FIG. 7 is a diagram showing an example of the operation of the continuity evaluation section 110a. In the example of FIG. 7, it is assumed that decoded blocks exist on the left side and above the image block to be encoded. The continuity evaluation unit 110a evaluates the continuity of the block boundary between the inter-predicted image (target image block) and the decoded block in order from top to bottom (or from bottom to top) pixel by pixel. For example, the continuity evaluation unit 110a calculates the continuity Cont of a block boundary spanning pixels p0 and q0 in FIG. 7 using the following equation (2).

Cont=｜p ₂ -2×p ₁ +p ₀ |+｜q ₂ -2×q ₁ +q ₀ | ...(2)

Note that the units and evaluation indexes for evaluating continuity are not limited to those shown in equation (2), and can be changed to other units and evaluation indexes as long as they are predefined in the system. For example, the unit of continuity evaluation may be the edge unit of the block boundary (continuity evaluation of the left boundary, continuity evaluation of the upper boundary), or the continuity evaluation index may be The pixel value difference (abs(p0-q0)) that straddles may be used.

The filter determination unit 110b determines whether to perform filter processing based on the prediction accuracy evaluation result input from the prediction accuracy evaluation unit 109 and the continuity evaluation result input from the continuity evaluation unit 110a. The filter determination unit 110b may perform the determination in units of pixels at the block boundary between the inter-predicted image and the decoded block, or may perform the determination in units of sides of the block boundary. The filter determination unit 110b may determine to change the strength of the filter (for example, the number of taps or frequency response of the filter) based on the prediction accuracy evaluation result and the continuity evaluation result. Note that when the prediction accuracy evaluation unit 109 evaluates the prediction accuracy on a pixel-by-pixel basis, the filter determination unit 110b calculates the average value of the estimated prediction accuracy R of each pixel in the edge region (target region of filter processing) of the inter-predicted image. It may also be used as a prediction accuracy evaluation result.

The filter determination unit 110b determines whether the prediction accuracy evaluated by the prediction accuracy evaluation unit 109 is less than or equal to the first threshold α, and the continuity evaluated by the continuity evaluation unit 110a is less than or equal to the second threshold β. , it is determined that filter processing is to be performed. On the other hand, if the prediction accuracy evaluated by the prediction accuracy evaluation unit 109 exceeds the first threshold α and/or the continuity evaluated by the continuity evaluation unit 110a exceeds the second threshold β, the filter determination unit 110b determines not to perform filter processing. Here, as the threshold values α and β, values fixed to the system may be set, or variable values may be calculated and set by a function using a quantization parameter (Qp). Furthermore, if the estimated prediction accuracy R is not normalized, the threshold α may be a variable value calculated and set by a function that uses the maximum value of the estimated prediction accuracy within the target image block.

The filter processing unit 110c performs filter processing on the end region of the inter-predicted image when the filter determination unit 110b determines that the filter processing is to be performed. As shown in FIG. 7, the filter processing unit 110c uses predicted pixels q0, q1, q2 included in the end region of the inter predicted image (block to be encoded) and adjacent decoded references input from the intra prediction unit 107. Pixels p0, p1, and p2 are used to perform correction according to equation (3) below, and corrected predicted pixels q'0, q'1, and q'2 are output.

Here, clip(x, max, min) indicates a process of clipping at max when x exceeds max and clipping at min when x is less than min. Further, tc in Equation (3) is an adjustment offset for processing used for clipping, and a variable value can be calculated and set by a function using a quantization parameter (Qp).

Although the predicted image correction unit 110 in the image encoding device 1 has been described, the predicted image correction unit 207 in the image decoding device 2 is configured similarly to the predicted image correction unit 110 in the image encoding device 1. Specifically, the predicted image correction section 207 in the image decoding device 2 includes a continuity evaluation section 207a, a filter determination section 207b, and a filter processing section 207c.

(6. Predicted image correction operation)
FIG. 8 is a diagram illustrating an example of a predicted image correction operation. Here, a predicted image correction operation in the image encoding device 1 will be described, but a similar predicted image correction operation is performed in the image decoding device 2 as well. The operation flow in FIG. 8 may be performed for each pixel at the block boundary between the inter-predicted image and the decoded block.

As shown in FIG. 8, in step S1, the prediction accuracy evaluation unit 109 determines whether the inter prediction unit 108 has a Evaluate prediction accuracy.

In step S2, the continuity evaluation unit 110a evaluates the continuity between the inter prediction image output by the inter prediction unit 108 and the decoded adjacent block adjacent to the inter prediction image.

In step S3, the filter determination unit 110b compares the prediction accuracy evaluated by the prediction accuracy evaluation unit 109 with a first threshold α, and compares the continuity evaluated by the continuity evaluation unit 110a with a second threshold β.

If the prediction accuracy is less than or equal to the first threshold α and the continuity is less than or equal to the second threshold β (step S3: YES), the filter determination unit 110b determines to perform filter processing. In such a case, in step S4, the filter processing unit 110c uses the decoded adjacent blocks (adjacent reference pixels) to correct the end regions of the inter-predicted image by filter processing.

On the other hand, if the prediction accuracy exceeds the first threshold α and/or the continuity exceeds the second threshold β (step S3: NO), the filter determination unit 110b determines not to perform filter processing. In such a case, the filter processing unit 110c does not perform filter processing.

(7. Summary of embodiments)
The image encoding device 1 and the image decoding device 2 according to this embodiment control filter processing based on the evaluation result of the prediction accuracy of inter prediction. As a result, when the prediction accuracy of the end region of the inter-predicted image is high, it is possible to control so as not to apply the filter processing to this end region, so that it is possible to prevent the precision of the predicted image from decreasing due to the filter processing. In other words, filter processing can be applied to the end region only when the prediction accuracy of the end region of the inter-predicted image is low, so that the filter processing can be performed appropriately.

Further, in this embodiment, the image encoding device 1 and the image decoding device 2 use a common operation to determine whether or not filter processing is applied. Specifically, the prediction accuracy of inter prediction is evaluated in each of the image encoding device 1 and the image decoding device 2, and based on the evaluation results, it is determined whether or not filter processing is applied. Thereby, there is no need to transmit a flag indicating whether filter processing is applied from the image encoding device 1 to the image decoding device 2, so that an increase in the amount of code due to the flag can be prevented.

(8. Modification example 1 of embodiment)
In the continuity evaluation and filter processing according to the embodiments described above, the intra prediction mode (intra prediction direction) applied to the decoded adjacent block may be considered. Thereby, the accuracy of continuity evaluation and filter processing can be improved. FIG. 9 is a diagram showing the configuration of the predicted image correction unit 110 according to this modification example. As shown in FIG. 9, the intra prediction mode applied to the decoded adjacent block is input from the intra prediction unit 107 to the continuity evaluation unit 110a and the filter processing unit 110c. The continuity evaluation unit 110a evaluates continuity according to the intra prediction mode applied to the decoded adjacent blocks. Furthermore, the filter processing unit 110c performs filter processing according to the intra prediction mode applied to the decoded adjacent block.

FIG. 10 is a diagram showing an example of the operation of the continuity evaluation section 110a and the filter processing section 110c according to this modification example. As shown in FIG. 10, the continuity evaluation unit 110a and the filter processing unit 110c perform continuity evaluation and filter processing on the end pixel q0 of the inter-predicted image corresponding to the current block to be encoded. In accordance with the intra prediction mode (intra prediction direction) applied to the block, p1, p0, q1, and q2 on a straight line passing through q0 are calculated by weighted average. Specifically, p1 and p0 on the straight line passing through q0 are calculated using the pixel values in the decoded adjacent block, and q1 and q2 on the straight line passing through q0 are calculated using the pixel values in the inter-predicted image. Calculate. The continuity evaluation unit 110a uses p1, p0, q1, and q2 calculated in this way to evaluate the continuity Cont according to the above equation (2). The filter processing unit 110c uses p1, p0, q1, and q2 calculated in this way to calculate q'0 according to the above equation (3).

(9. Modification example 2 of embodiment)
In the embodiment described above, an example has been described in which the predicted image correction unit 110 performs filter processing on the end region of the inter-predicted image using decoded adjacent blocks as correction processing on the inter-predicted image.

Furthermore, the correction processing is not limited to such filter processing, and weighted averaging processing of inter-predicted images and intra-predicted images may be used as the correction processing.

In this modification example, the intra prediction unit 107 and predicted image correction unit 110 of the image encoding device 1 shown in FIG. 1 perform correction processing on the inter predicted image using decoded adjacent blocks adjacent to the target image block. It constitutes a correction section.

The intra prediction unit 107 generates an intra predicted image corresponding to the target image block using the decoded adjacent blocks. For example, the intra prediction unit 107 generates an intra prediction image from decoded adjacent blocks using a predetermined intra prediction mode. The predicted image correction unit 110 performs weighted averaging processing of the inter predicted image input from the inter prediction unit 108 and the intra predicted image input from the intra prediction unit 107 as correction processing. Note that the transmitting side may select one prediction mode from among a plurality of intra prediction mode candidates and generate an intra prediction image. At that time, the entropy encoding unit 103 entropy encodes the intra prediction mode flag indicating which intra prediction mode was used to generate the intra predicted image on the transmitting side.

Similarly, the intra prediction unit 204 and the predicted image correction unit 207 of the image decoding device 2 shown in FIG. Configure. The intra prediction unit 204 generates an intra predicted image corresponding to the target image block using the decoded adjacent blocks.

For example, the intra prediction unit 204 generates an intra prediction image from decoded adjacent blocks using a predetermined intra prediction mode. The predicted image correction unit 207 performs weighted averaging processing of the inter predicted image input from the inter prediction unit 205 and the intra predicted image input from the intra prediction unit 204 as correction processing. Note that the intra prediction mode used for generating the intra predicted image may be determined based on the intra prediction mode flag sent from the transmitting side.

When applying the configuration of this modified example, "filter processing" in the embodiment described above may be replaced with "weighted average processing." Furthermore, in addition to determining whether or not to perform such weighted averaging processing, the weight in weighted averaging processing may be adjusted based on the evaluation result of prediction accuracy and the evaluation result of continuity.

(10. Other embodiments)
In the embodiments described above, bi-prediction was mainly explained as prediction using a plurality of reference images. However, it is also possible to apply a technique called intra block copy as prediction using multiple reference images. In intra block copying, a reference image within the same frame as the current frame is used to predict the target image block of the current frame.

It may be provided by a program that causes a computer to execute each process performed by the image encoding device 1 and a program that causes a computer to execute each process performed by the image decoding device 2. Moreover, the program may be recorded on a computer-readable medium. Computer-readable media allow programs to be installed on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM. Further, the image encoding device 1 may be configured as a semiconductor integrated circuit (chip set, SoC) by integrating circuits that execute each process performed by the image encoding device 1. Similarly, the image decoding device 2 may be configured as a semiconductor integrated circuit (chip set, SoC) by integrating circuits that execute each process performed by the image decoding device 2.

Although the embodiments have been described above in detail with reference to the drawings, the specific configuration is not limited to that described above, and various design changes can be made without departing from the scope of the invention.

Note that the entire contents of Japanese Patent Application No. 2018-72451 (filed on April 4, 2018) are incorporated into the specification of the present application by reference.

Claims (4)

an inter prediction unit that generates an inter prediction block by predicting a block of a target image obtained by dividing the image in units of frames by inter prediction;
a correction unit that performs correction processing on the inter prediction block;
comprising a calculation section ;
Only when the inter prediction unit performs the inter prediction using blocks of a plurality of reference images, the calculation unit calculates the statistics of errors between blocks of the plurality of reference images for a portion of the blocks. Calculate each image part consisting of multiple pixels using
The correction unit performs a correction process on only the image portion in which the error statistics satisfy a predetermined threshold condition among the blocks of the predicted image. An image encoding device comprising the predicted image correction device according to claim 1 . An image decoding device comprising the predicted image correction device according to claim 1 . A program that causes a computer to function as the predictive image correction device according to claim 1 .

Images